Commit
•
af3c6ae
1
Parent(s):
ebf87dd
End of training
Browse files- all_results.json +13 -0
- eval_results.json +8 -0
- runs/Jun28_23-30-42_51c7928aca28/events.out.tfevents.1719629219.51c7928aca28.34.3 +3 -0
- train_results.json +8 -0
- trainer_state.json +2092 -0
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.973285841495994,
|
| 3 |
+
"eval_accuracy": 0.9900356268091739,
|
| 4 |
+
"eval_loss": 0.04398982226848602,
|
| 5 |
+
"eval_runtime": 168.2023,
|
| 6 |
+
"eval_samples_per_second": 106.8,
|
| 7 |
+
"eval_steps_per_second": 1.671,
|
| 8 |
+
"total_flos": 1.7813294264095949e+19,
|
| 9 |
+
"train_loss": 0.08161249804892577,
|
| 10 |
+
"train_runtime": 11582.5203,
|
| 11 |
+
"train_samples_per_second": 62.035,
|
| 12 |
+
"train_steps_per_second": 0.242
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.973285841495994,
|
| 3 |
+
"eval_accuracy": 0.9900356268091739,
|
| 4 |
+
"eval_loss": 0.04398982226848602,
|
| 5 |
+
"eval_runtime": 168.2023,
|
| 6 |
+
"eval_samples_per_second": 106.8,
|
| 7 |
+
"eval_steps_per_second": 1.671
|
| 8 |
+
}
|
runs/Jun28_23-30-42_51c7928aca28/events.out.tfevents.1719629219.51c7928aca28.34.3
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:036026e503284fcd1623f018e5dc03a3766710e9c21d1460ea330cd3bf888f51
|
| 3 |
+
size 411
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.973285841495994,
|
| 3 |
+
"total_flos": 1.7813294264095949e+19,
|
| 4 |
+
"train_loss": 0.08161249804892577,
|
| 5 |
+
"train_runtime": 11582.5203,
|
| 6 |
+
"train_samples_per_second": 62.035,
|
| 7 |
+
"train_steps_per_second": 0.242
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2092 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.9900356268091739,
|
| 3 |
+
"best_model_checkpoint": "Garbage-Classification-SWIN-Transformer/checkpoint-2800",
|
| 4 |
+
"epoch": 9.973285841495994,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2800,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.03561887800534283,
|
| 13 |
+
"grad_norm": 3.4222159385681152,
|
| 14 |
+
"learning_rate": 1.7857142857142857e-06,
|
| 15 |
+
"loss": 1.4083,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.07123775601068566,
|
| 20 |
+
"grad_norm": 3.2973814010620117,
|
| 21 |
+
"learning_rate": 3.5714285714285714e-06,
|
| 22 |
+
"loss": 1.3949,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.10685663401602849,
|
| 27 |
+
"grad_norm": 3.3026835918426514,
|
| 28 |
+
"learning_rate": 5.357142857142857e-06,
|
| 29 |
+
"loss": 1.3661,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.14247551202137132,
|
| 34 |
+
"grad_norm": 4.8538289070129395,
|
| 35 |
+
"learning_rate": 7.142857142857143e-06,
|
| 36 |
+
"loss": 1.3301,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.17809439002671415,
|
| 41 |
+
"grad_norm": 6.877591609954834,
|
| 42 |
+
"learning_rate": 8.92857142857143e-06,
|
| 43 |
+
"loss": 1.2576,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.21371326803205698,
|
| 48 |
+
"grad_norm": 5.944957256317139,
|
| 49 |
+
"learning_rate": 1.0714285714285714e-05,
|
| 50 |
+
"loss": 1.1654,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.2493321460373998,
|
| 55 |
+
"grad_norm": 6.024969100952148,
|
| 56 |
+
"learning_rate": 1.25e-05,
|
| 57 |
+
"loss": 1.0324,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.28495102404274264,
|
| 62 |
+
"grad_norm": 6.543083667755127,
|
| 63 |
+
"learning_rate": 1.4285714285714285e-05,
|
| 64 |
+
"loss": 0.8479,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.3205699020480855,
|
| 69 |
+
"grad_norm": 5.732212543487549,
|
| 70 |
+
"learning_rate": 1.6071428571428572e-05,
|
| 71 |
+
"loss": 0.7139,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.3561887800534283,
|
| 76 |
+
"grad_norm": 6.9801459312438965,
|
| 77 |
+
"learning_rate": 1.785714285714286e-05,
|
| 78 |
+
"loss": 0.5947,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.39180765805877116,
|
| 83 |
+
"grad_norm": 7.786228656768799,
|
| 84 |
+
"learning_rate": 1.9642857142857145e-05,
|
| 85 |
+
"loss": 0.5308,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.42742653606411396,
|
| 90 |
+
"grad_norm": 7.892849445343018,
|
| 91 |
+
"learning_rate": 2.1428571428571428e-05,
|
| 92 |
+
"loss": 0.4402,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.4630454140694568,
|
| 97 |
+
"grad_norm": 6.560166835784912,
|
| 98 |
+
"learning_rate": 2.3214285714285715e-05,
|
| 99 |
+
"loss": 0.4023,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.4986642920747996,
|
| 104 |
+
"grad_norm": 7.141245365142822,
|
| 105 |
+
"learning_rate": 2.5e-05,
|
| 106 |
+
"loss": 0.3745,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.5342831700801425,
|
| 111 |
+
"grad_norm": 4.8248419761657715,
|
| 112 |
+
"learning_rate": 2.6785714285714288e-05,
|
| 113 |
+
"loss": 0.3398,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.5699020480854853,
|
| 118 |
+
"grad_norm": 5.556511878967285,
|
| 119 |
+
"learning_rate": 2.857142857142857e-05,
|
| 120 |
+
"loss": 0.2968,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.6055209260908282,
|
| 125 |
+
"grad_norm": 7.37336540222168,
|
| 126 |
+
"learning_rate": 3.0357142857142857e-05,
|
| 127 |
+
"loss": 0.2907,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.641139804096171,
|
| 132 |
+
"grad_norm": 6.2109503746032715,
|
| 133 |
+
"learning_rate": 3.2142857142857144e-05,
|
| 134 |
+
"loss": 0.2644,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.6767586821015138,
|
| 139 |
+
"grad_norm": 5.759757995605469,
|
| 140 |
+
"learning_rate": 3.392857142857143e-05,
|
| 141 |
+
"loss": 0.2918,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.7123775601068566,
|
| 146 |
+
"grad_norm": 3.9079067707061768,
|
| 147 |
+
"learning_rate": 3.571428571428572e-05,
|
| 148 |
+
"loss": 0.251,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.7479964381121995,
|
| 153 |
+
"grad_norm": 4.728653430938721,
|
| 154 |
+
"learning_rate": 3.7500000000000003e-05,
|
| 155 |
+
"loss": 0.255,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.7836153161175423,
|
| 160 |
+
"grad_norm": 5.230493545532227,
|
| 161 |
+
"learning_rate": 3.928571428571429e-05,
|
| 162 |
+
"loss": 0.2118,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.8192341941228851,
|
| 167 |
+
"grad_norm": 5.2034101486206055,
|
| 168 |
+
"learning_rate": 4.107142857142857e-05,
|
| 169 |
+
"loss": 0.2045,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.8548530721282279,
|
| 174 |
+
"grad_norm": 5.750576972961426,
|
| 175 |
+
"learning_rate": 4.2857142857142856e-05,
|
| 176 |
+
"loss": 0.2076,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.8904719501335708,
|
| 181 |
+
"grad_norm": 4.098123550415039,
|
| 182 |
+
"learning_rate": 4.464285714285715e-05,
|
| 183 |
+
"loss": 0.2077,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.9260908281389136,
|
| 188 |
+
"grad_norm": 6.287376880645752,
|
| 189 |
+
"learning_rate": 4.642857142857143e-05,
|
| 190 |
+
"loss": 0.1989,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.9617097061442564,
|
| 195 |
+
"grad_norm": 5.270883083343506,
|
| 196 |
+
"learning_rate": 4.8214285714285716e-05,
|
| 197 |
+
"loss": 0.1782,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.9973285841495992,
|
| 202 |
+
"grad_norm": 9.215749740600586,
|
| 203 |
+
"learning_rate": 5e-05,
|
| 204 |
+
"loss": 0.1969,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.9973285841495992,
|
| 209 |
+
"eval_accuracy": 0.9409374304163883,
|
| 210 |
+
"eval_loss": 0.17398740351200104,
|
| 211 |
+
"eval_runtime": 169.1651,
|
| 212 |
+
"eval_samples_per_second": 106.192,
|
| 213 |
+
"eval_steps_per_second": 1.661,
|
| 214 |
+
"step": 280
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 1.0329474621549422,
|
| 218 |
+
"grad_norm": 3.8531081676483154,
|
| 219 |
+
"learning_rate": 4.9801587301587306e-05,
|
| 220 |
+
"loss": 0.1794,
|
| 221 |
+
"step": 290
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 1.068566340160285,
|
| 225 |
+
"grad_norm": 5.727213382720947,
|
| 226 |
+
"learning_rate": 4.960317460317461e-05,
|
| 227 |
+
"loss": 0.1792,
|
| 228 |
+
"step": 300
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 1.1041852181656278,
|
| 232 |
+
"grad_norm": 5.829253673553467,
|
| 233 |
+
"learning_rate": 4.940476190476191e-05,
|
| 234 |
+
"loss": 0.1445,
|
| 235 |
+
"step": 310
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 1.1398040961709706,
|
| 239 |
+
"grad_norm": 5.027329921722412,
|
| 240 |
+
"learning_rate": 4.9206349206349204e-05,
|
| 241 |
+
"loss": 0.1282,
|
| 242 |
+
"step": 320
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 1.1754229741763134,
|
| 246 |
+
"grad_norm": 3.377734422683716,
|
| 247 |
+
"learning_rate": 4.900793650793651e-05,
|
| 248 |
+
"loss": 0.1346,
|
| 249 |
+
"step": 330
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 1.2110418521816562,
|
| 253 |
+
"grad_norm": 4.863683700561523,
|
| 254 |
+
"learning_rate": 4.880952380952381e-05,
|
| 255 |
+
"loss": 0.1141,
|
| 256 |
+
"step": 340
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 1.2466607301869992,
|
| 260 |
+
"grad_norm": 6.36651086807251,
|
| 261 |
+
"learning_rate": 4.8611111111111115e-05,
|
| 262 |
+
"loss": 0.129,
|
| 263 |
+
"step": 350
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 1.282279608192342,
|
| 267 |
+
"grad_norm": 6.058999061584473,
|
| 268 |
+
"learning_rate": 4.841269841269841e-05,
|
| 269 |
+
"loss": 0.1304,
|
| 270 |
+
"step": 360
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 1.3178984861976848,
|
| 274 |
+
"grad_norm": 3.276484251022339,
|
| 275 |
+
"learning_rate": 4.8214285714285716e-05,
|
| 276 |
+
"loss": 0.1107,
|
| 277 |
+
"step": 370
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 1.3535173642030276,
|
| 281 |
+
"grad_norm": 4.594018936157227,
|
| 282 |
+
"learning_rate": 4.801587301587302e-05,
|
| 283 |
+
"loss": 0.1185,
|
| 284 |
+
"step": 380
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 1.3891362422083704,
|
| 288 |
+
"grad_norm": 2.975635290145874,
|
| 289 |
+
"learning_rate": 4.781746031746032e-05,
|
| 290 |
+
"loss": 0.1287,
|
| 291 |
+
"step": 390
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 1.4247551202137132,
|
| 295 |
+
"grad_norm": 3.8723185062408447,
|
| 296 |
+
"learning_rate": 4.761904761904762e-05,
|
| 297 |
+
"loss": 0.1207,
|
| 298 |
+
"step": 400
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 1.460373998219056,
|
| 302 |
+
"grad_norm": 2.9883522987365723,
|
| 303 |
+
"learning_rate": 4.7420634920634924e-05,
|
| 304 |
+
"loss": 0.1012,
|
| 305 |
+
"step": 410
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 1.495992876224399,
|
| 309 |
+
"grad_norm": 4.591331481933594,
|
| 310 |
+
"learning_rate": 4.722222222222222e-05,
|
| 311 |
+
"loss": 0.1092,
|
| 312 |
+
"step": 420
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 1.5316117542297416,
|
| 316 |
+
"grad_norm": 3.980555295944214,
|
| 317 |
+
"learning_rate": 4.7023809523809525e-05,
|
| 318 |
+
"loss": 0.1025,
|
| 319 |
+
"step": 430
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 1.5672306322350846,
|
| 323 |
+
"grad_norm": 4.750612258911133,
|
| 324 |
+
"learning_rate": 4.682539682539683e-05,
|
| 325 |
+
"loss": 0.1047,
|
| 326 |
+
"step": 440
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.6028495102404274,
|
| 330 |
+
"grad_norm": 3.2454400062561035,
|
| 331 |
+
"learning_rate": 4.662698412698413e-05,
|
| 332 |
+
"loss": 0.0914,
|
| 333 |
+
"step": 450
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 1.6384683882457702,
|
| 337 |
+
"grad_norm": 11.585031509399414,
|
| 338 |
+
"learning_rate": 4.642857142857143e-05,
|
| 339 |
+
"loss": 0.0827,
|
| 340 |
+
"step": 460
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 1.674087266251113,
|
| 344 |
+
"grad_norm": 2.573432207107544,
|
| 345 |
+
"learning_rate": 4.623015873015873e-05,
|
| 346 |
+
"loss": 0.0866,
|
| 347 |
+
"step": 470
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 1.7097061442564558,
|
| 351 |
+
"grad_norm": 3.991711378097534,
|
| 352 |
+
"learning_rate": 4.603174603174603e-05,
|
| 353 |
+
"loss": 0.088,
|
| 354 |
+
"step": 480
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.7453250222617989,
|
| 358 |
+
"grad_norm": 4.302712440490723,
|
| 359 |
+
"learning_rate": 4.5833333333333334e-05,
|
| 360 |
+
"loss": 0.0985,
|
| 361 |
+
"step": 490
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 1.7809439002671414,
|
| 365 |
+
"grad_norm": 3.247912883758545,
|
| 366 |
+
"learning_rate": 4.563492063492064e-05,
|
| 367 |
+
"loss": 0.1015,
|
| 368 |
+
"step": 500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.8165627782724845,
|
| 372 |
+
"grad_norm": 3.324694871902466,
|
| 373 |
+
"learning_rate": 4.543650793650794e-05,
|
| 374 |
+
"loss": 0.101,
|
| 375 |
+
"step": 510
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 1.8521816562778273,
|
| 379 |
+
"grad_norm": 5.875315189361572,
|
| 380 |
+
"learning_rate": 4.523809523809524e-05,
|
| 381 |
+
"loss": 0.0908,
|
| 382 |
+
"step": 520
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 1.88780053428317,
|
| 386 |
+
"grad_norm": 2.6438138484954834,
|
| 387 |
+
"learning_rate": 4.503968253968254e-05,
|
| 388 |
+
"loss": 0.0979,
|
| 389 |
+
"step": 530
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.9234194122885129,
|
| 393 |
+
"grad_norm": 2.6687426567077637,
|
| 394 |
+
"learning_rate": 4.4841269841269846e-05,
|
| 395 |
+
"loss": 0.0834,
|
| 396 |
+
"step": 540
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 1.9590382902938557,
|
| 400 |
+
"grad_norm": 3.7509348392486572,
|
| 401 |
+
"learning_rate": 4.464285714285715e-05,
|
| 402 |
+
"loss": 0.0826,
|
| 403 |
+
"step": 550
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 1.9946571682991987,
|
| 407 |
+
"grad_norm": 3.0679826736450195,
|
| 408 |
+
"learning_rate": 4.4444444444444447e-05,
|
| 409 |
+
"loss": 0.1014,
|
| 410 |
+
"step": 560
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.9982190560997328,
|
| 414 |
+
"eval_accuracy": 0.9755065686929414,
|
| 415 |
+
"eval_loss": 0.07521907240152359,
|
| 416 |
+
"eval_runtime": 167.9798,
|
| 417 |
+
"eval_samples_per_second": 106.941,
|
| 418 |
+
"eval_steps_per_second": 1.673,
|
| 419 |
+
"step": 561
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 2.0302760463045413,
|
| 423 |
+
"grad_norm": 4.248985290527344,
|
| 424 |
+
"learning_rate": 4.4246031746031744e-05,
|
| 425 |
+
"loss": 0.0633,
|
| 426 |
+
"step": 570
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 2.0658949243098843,
|
| 430 |
+
"grad_norm": 3.44899320602417,
|
| 431 |
+
"learning_rate": 4.404761904761905e-05,
|
| 432 |
+
"loss": 0.0665,
|
| 433 |
+
"step": 580
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 2.101513802315227,
|
| 437 |
+
"grad_norm": 4.588482856750488,
|
| 438 |
+
"learning_rate": 4.384920634920635e-05,
|
| 439 |
+
"loss": 0.0503,
|
| 440 |
+
"step": 590
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 2.13713268032057,
|
| 444 |
+
"grad_norm": 3.4909369945526123,
|
| 445 |
+
"learning_rate": 4.3650793650793655e-05,
|
| 446 |
+
"loss": 0.0513,
|
| 447 |
+
"step": 600
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 2.172751558325913,
|
| 451 |
+
"grad_norm": 2.8658242225646973,
|
| 452 |
+
"learning_rate": 4.345238095238096e-05,
|
| 453 |
+
"loss": 0.0552,
|
| 454 |
+
"step": 610
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 2.2083704363312555,
|
| 458 |
+
"grad_norm": 3.1010851860046387,
|
| 459 |
+
"learning_rate": 4.3253968253968256e-05,
|
| 460 |
+
"loss": 0.0468,
|
| 461 |
+
"step": 620
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 2.2439893143365985,
|
| 465 |
+
"grad_norm": 1.9300130605697632,
|
| 466 |
+
"learning_rate": 4.305555555555556e-05,
|
| 467 |
+
"loss": 0.0551,
|
| 468 |
+
"step": 630
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 2.279608192341941,
|
| 472 |
+
"grad_norm": 2.8282604217529297,
|
| 473 |
+
"learning_rate": 4.2857142857142856e-05,
|
| 474 |
+
"loss": 0.0525,
|
| 475 |
+
"step": 640
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 2.315227070347284,
|
| 479 |
+
"grad_norm": 4.43895149230957,
|
| 480 |
+
"learning_rate": 4.265873015873016e-05,
|
| 481 |
+
"loss": 0.0492,
|
| 482 |
+
"step": 650
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 2.3508459483526267,
|
| 486 |
+
"grad_norm": 4.581601142883301,
|
| 487 |
+
"learning_rate": 4.2460317460317464e-05,
|
| 488 |
+
"loss": 0.0506,
|
| 489 |
+
"step": 660
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 2.3864648263579697,
|
| 493 |
+
"grad_norm": 4.677193641662598,
|
| 494 |
+
"learning_rate": 4.226190476190476e-05,
|
| 495 |
+
"loss": 0.0404,
|
| 496 |
+
"step": 670
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 2.4220837043633123,
|
| 500 |
+
"grad_norm": 2.8398616313934326,
|
| 501 |
+
"learning_rate": 4.2063492063492065e-05,
|
| 502 |
+
"loss": 0.0438,
|
| 503 |
+
"step": 680
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 2.4577025823686554,
|
| 507 |
+
"grad_norm": 5.424591064453125,
|
| 508 |
+
"learning_rate": 4.186507936507937e-05,
|
| 509 |
+
"loss": 0.0544,
|
| 510 |
+
"step": 690
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 2.4933214603739984,
|
| 514 |
+
"grad_norm": 4.595205307006836,
|
| 515 |
+
"learning_rate": 4.166666666666667e-05,
|
| 516 |
+
"loss": 0.0557,
|
| 517 |
+
"step": 700
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 2.528940338379341,
|
| 521 |
+
"grad_norm": 3.7095601558685303,
|
| 522 |
+
"learning_rate": 4.1468253968253976e-05,
|
| 523 |
+
"loss": 0.0505,
|
| 524 |
+
"step": 710
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 2.564559216384684,
|
| 528 |
+
"grad_norm": 2.084617853164673,
|
| 529 |
+
"learning_rate": 4.126984126984127e-05,
|
| 530 |
+
"loss": 0.037,
|
| 531 |
+
"step": 720
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 2.6001780943900266,
|
| 535 |
+
"grad_norm": 3.442814350128174,
|
| 536 |
+
"learning_rate": 4.107142857142857e-05,
|
| 537 |
+
"loss": 0.0399,
|
| 538 |
+
"step": 730
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 2.6357969723953696,
|
| 542 |
+
"grad_norm": 3.6135501861572266,
|
| 543 |
+
"learning_rate": 4.0873015873015874e-05,
|
| 544 |
+
"loss": 0.0498,
|
| 545 |
+
"step": 740
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 2.671415850400712,
|
| 549 |
+
"grad_norm": 3.422394275665283,
|
| 550 |
+
"learning_rate": 4.067460317460318e-05,
|
| 551 |
+
"loss": 0.0526,
|
| 552 |
+
"step": 750
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 2.707034728406055,
|
| 556 |
+
"grad_norm": 4.467539310455322,
|
| 557 |
+
"learning_rate": 4.047619047619048e-05,
|
| 558 |
+
"loss": 0.0392,
|
| 559 |
+
"step": 760
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 2.742653606411398,
|
| 563 |
+
"grad_norm": 2.9515881538391113,
|
| 564 |
+
"learning_rate": 4.027777777777778e-05,
|
| 565 |
+
"loss": 0.0468,
|
| 566 |
+
"step": 770
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 2.778272484416741,
|
| 570 |
+
"grad_norm": 4.296915054321289,
|
| 571 |
+
"learning_rate": 4.007936507936508e-05,
|
| 572 |
+
"loss": 0.0534,
|
| 573 |
+
"step": 780
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 2.813891362422084,
|
| 577 |
+
"grad_norm": 2.723884344100952,
|
| 578 |
+
"learning_rate": 3.9880952380952386e-05,
|
| 579 |
+
"loss": 0.048,
|
| 580 |
+
"step": 790
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 2.8495102404274264,
|
| 584 |
+
"grad_norm": 1.9511892795562744,
|
| 585 |
+
"learning_rate": 3.968253968253968e-05,
|
| 586 |
+
"loss": 0.0429,
|
| 587 |
+
"step": 800
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 2.8851291184327694,
|
| 591 |
+
"grad_norm": 2.8428244590759277,
|
| 592 |
+
"learning_rate": 3.9484126984126986e-05,
|
| 593 |
+
"loss": 0.0386,
|
| 594 |
+
"step": 810
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 2.920747996438112,
|
| 598 |
+
"grad_norm": 5.256746768951416,
|
| 599 |
+
"learning_rate": 3.928571428571429e-05,
|
| 600 |
+
"loss": 0.0421,
|
| 601 |
+
"step": 820
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 2.956366874443455,
|
| 605 |
+
"grad_norm": 2.277379035949707,
|
| 606 |
+
"learning_rate": 3.908730158730159e-05,
|
| 607 |
+
"loss": 0.0521,
|
| 608 |
+
"step": 830
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 2.991985752448798,
|
| 612 |
+
"grad_norm": 2.638746500015259,
|
| 613 |
+
"learning_rate": 3.888888888888889e-05,
|
| 614 |
+
"loss": 0.0333,
|
| 615 |
+
"step": 840
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 2.9991095280498663,
|
| 619 |
+
"eval_accuracy": 0.9824092629703852,
|
| 620 |
+
"eval_loss": 0.05514274910092354,
|
| 621 |
+
"eval_runtime": 167.4573,
|
| 622 |
+
"eval_samples_per_second": 107.275,
|
| 623 |
+
"eval_steps_per_second": 1.678,
|
| 624 |
+
"step": 842
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 3.0276046304541406,
|
| 628 |
+
"grad_norm": 2.1443052291870117,
|
| 629 |
+
"learning_rate": 3.8690476190476195e-05,
|
| 630 |
+
"loss": 0.0221,
|
| 631 |
+
"step": 850
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"epoch": 3.0632235084594837,
|
| 635 |
+
"grad_norm": 2.5836849212646484,
|
| 636 |
+
"learning_rate": 3.84920634920635e-05,
|
| 637 |
+
"loss": 0.0258,
|
| 638 |
+
"step": 860
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 3.0988423864648262,
|
| 642 |
+
"grad_norm": 4.255194187164307,
|
| 643 |
+
"learning_rate": 3.8293650793650795e-05,
|
| 644 |
+
"loss": 0.0269,
|
| 645 |
+
"step": 870
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 3.1344612644701693,
|
| 649 |
+
"grad_norm": 1.1687568426132202,
|
| 650 |
+
"learning_rate": 3.809523809523809e-05,
|
| 651 |
+
"loss": 0.0229,
|
| 652 |
+
"step": 880
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 3.170080142475512,
|
| 656 |
+
"grad_norm": 3.9044461250305176,
|
| 657 |
+
"learning_rate": 3.7896825396825396e-05,
|
| 658 |
+
"loss": 0.0241,
|
| 659 |
+
"step": 890
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 3.205699020480855,
|
| 663 |
+
"grad_norm": 3.447392225265503,
|
| 664 |
+
"learning_rate": 3.76984126984127e-05,
|
| 665 |
+
"loss": 0.0258,
|
| 666 |
+
"step": 900
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 3.241317898486198,
|
| 670 |
+
"grad_norm": 1.6584926843643188,
|
| 671 |
+
"learning_rate": 3.7500000000000003e-05,
|
| 672 |
+
"loss": 0.0221,
|
| 673 |
+
"step": 910
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 3.2769367764915405,
|
| 677 |
+
"grad_norm": 6.36900520324707,
|
| 678 |
+
"learning_rate": 3.730158730158731e-05,
|
| 679 |
+
"loss": 0.0272,
|
| 680 |
+
"step": 920
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 3.3125556544968835,
|
| 684 |
+
"grad_norm": 1.01226007938385,
|
| 685 |
+
"learning_rate": 3.7103174603174604e-05,
|
| 686 |
+
"loss": 0.0242,
|
| 687 |
+
"step": 930
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 3.348174532502226,
|
| 691 |
+
"grad_norm": 0.7302751541137695,
|
| 692 |
+
"learning_rate": 3.690476190476191e-05,
|
| 693 |
+
"loss": 0.0276,
|
| 694 |
+
"step": 940
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 3.383793410507569,
|
| 698 |
+
"grad_norm": 5.937874794006348,
|
| 699 |
+
"learning_rate": 3.6706349206349205e-05,
|
| 700 |
+
"loss": 0.03,
|
| 701 |
+
"step": 950
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 3.4194122885129117,
|
| 705 |
+
"grad_norm": 3.688383102416992,
|
| 706 |
+
"learning_rate": 3.650793650793651e-05,
|
| 707 |
+
"loss": 0.0239,
|
| 708 |
+
"step": 960
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 3.4550311665182547,
|
| 712 |
+
"grad_norm": 8.206720352172852,
|
| 713 |
+
"learning_rate": 3.630952380952381e-05,
|
| 714 |
+
"loss": 0.038,
|
| 715 |
+
"step": 970
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 3.4906500445235977,
|
| 719 |
+
"grad_norm": 1.0089924335479736,
|
| 720 |
+
"learning_rate": 3.611111111111111e-05,
|
| 721 |
+
"loss": 0.0275,
|
| 722 |
+
"step": 980
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 3.5262689225289403,
|
| 726 |
+
"grad_norm": 4.124368190765381,
|
| 727 |
+
"learning_rate": 3.591269841269841e-05,
|
| 728 |
+
"loss": 0.0383,
|
| 729 |
+
"step": 990
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 3.5618878005342833,
|
| 733 |
+
"grad_norm": 4.853917598724365,
|
| 734 |
+
"learning_rate": 3.571428571428572e-05,
|
| 735 |
+
"loss": 0.0183,
|
| 736 |
+
"step": 1000
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 3.597506678539626,
|
| 740 |
+
"grad_norm": 0.8174078464508057,
|
| 741 |
+
"learning_rate": 3.551587301587302e-05,
|
| 742 |
+
"loss": 0.0176,
|
| 743 |
+
"step": 1010
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 3.633125556544969,
|
| 747 |
+
"grad_norm": 2.9447782039642334,
|
| 748 |
+
"learning_rate": 3.5317460317460324e-05,
|
| 749 |
+
"loss": 0.0294,
|
| 750 |
+
"step": 1020
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 3.6687444345503115,
|
| 754 |
+
"grad_norm": 2.926154375076294,
|
| 755 |
+
"learning_rate": 3.511904761904762e-05,
|
| 756 |
+
"loss": 0.0257,
|
| 757 |
+
"step": 1030
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 3.7043633125556545,
|
| 761 |
+
"grad_norm": 2.566096305847168,
|
| 762 |
+
"learning_rate": 3.492063492063492e-05,
|
| 763 |
+
"loss": 0.0172,
|
| 764 |
+
"step": 1040
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 3.7399821905609976,
|
| 768 |
+
"grad_norm": 3.302891492843628,
|
| 769 |
+
"learning_rate": 3.472222222222222e-05,
|
| 770 |
+
"loss": 0.0314,
|
| 771 |
+
"step": 1050
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 3.77560106856634,
|
| 775 |
+
"grad_norm": 1.366655945777893,
|
| 776 |
+
"learning_rate": 3.4523809523809526e-05,
|
| 777 |
+
"loss": 0.0243,
|
| 778 |
+
"step": 1060
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 3.8112199465716827,
|
| 782 |
+
"grad_norm": 1.3453521728515625,
|
| 783 |
+
"learning_rate": 3.432539682539683e-05,
|
| 784 |
+
"loss": 0.0166,
|
| 785 |
+
"step": 1070
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 3.8468388245770258,
|
| 789 |
+
"grad_norm": 3.0241684913635254,
|
| 790 |
+
"learning_rate": 3.412698412698413e-05,
|
| 791 |
+
"loss": 0.0341,
|
| 792 |
+
"step": 1080
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 3.8824577025823688,
|
| 796 |
+
"grad_norm": 4.319317817687988,
|
| 797 |
+
"learning_rate": 3.392857142857143e-05,
|
| 798 |
+
"loss": 0.0199,
|
| 799 |
+
"step": 1090
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 3.9180765805877114,
|
| 803 |
+
"grad_norm": 0.412803053855896,
|
| 804 |
+
"learning_rate": 3.3730158730158734e-05,
|
| 805 |
+
"loss": 0.0282,
|
| 806 |
+
"step": 1100
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 3.9536954585930544,
|
| 810 |
+
"grad_norm": 2.4635729789733887,
|
| 811 |
+
"learning_rate": 3.353174603174603e-05,
|
| 812 |
+
"loss": 0.0201,
|
| 813 |
+
"step": 1110
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 3.9893143365983974,
|
| 817 |
+
"grad_norm": 3.8697004318237305,
|
| 818 |
+
"learning_rate": 3.3333333333333335e-05,
|
| 819 |
+
"loss": 0.0332,
|
| 820 |
+
"step": 1120
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 4.0,
|
| 824 |
+
"eval_accuracy": 0.9844689378757515,
|
| 825 |
+
"eval_loss": 0.05261413753032684,
|
| 826 |
+
"eval_runtime": 166.8993,
|
| 827 |
+
"eval_samples_per_second": 107.634,
|
| 828 |
+
"eval_steps_per_second": 1.684,
|
| 829 |
+
"step": 1123
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 4.02493321460374,
|
| 833 |
+
"grad_norm": 0.7120935916900635,
|
| 834 |
+
"learning_rate": 3.313492063492064e-05,
|
| 835 |
+
"loss": 0.0137,
|
| 836 |
+
"step": 1130
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 4.060552092609083,
|
| 840 |
+
"grad_norm": 4.946304798126221,
|
| 841 |
+
"learning_rate": 3.2936507936507936e-05,
|
| 842 |
+
"loss": 0.0199,
|
| 843 |
+
"step": 1140
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 4.096170970614426,
|
| 847 |
+
"grad_norm": 3.8117804527282715,
|
| 848 |
+
"learning_rate": 3.273809523809524e-05,
|
| 849 |
+
"loss": 0.0177,
|
| 850 |
+
"step": 1150
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 4.131789848619769,
|
| 854 |
+
"grad_norm": 1.8831413984298706,
|
| 855 |
+
"learning_rate": 3.253968253968254e-05,
|
| 856 |
+
"loss": 0.0197,
|
| 857 |
+
"step": 1160
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 4.167408726625111,
|
| 861 |
+
"grad_norm": 0.7759003639221191,
|
| 862 |
+
"learning_rate": 3.234126984126985e-05,
|
| 863 |
+
"loss": 0.0151,
|
| 864 |
+
"step": 1170
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 4.203027604630454,
|
| 868 |
+
"grad_norm": 4.501601696014404,
|
| 869 |
+
"learning_rate": 3.2142857142857144e-05,
|
| 870 |
+
"loss": 0.0148,
|
| 871 |
+
"step": 1180
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 4.238646482635797,
|
| 875 |
+
"grad_norm": 2.070884943008423,
|
| 876 |
+
"learning_rate": 3.194444444444444e-05,
|
| 877 |
+
"loss": 0.0176,
|
| 878 |
+
"step": 1190
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 4.27426536064114,
|
| 882 |
+
"grad_norm": 2.1252646446228027,
|
| 883 |
+
"learning_rate": 3.1746031746031745e-05,
|
| 884 |
+
"loss": 0.0141,
|
| 885 |
+
"step": 1200
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 4.309884238646482,
|
| 889 |
+
"grad_norm": 2.5485196113586426,
|
| 890 |
+
"learning_rate": 3.154761904761905e-05,
|
| 891 |
+
"loss": 0.017,
|
| 892 |
+
"step": 1210
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 4.345503116651826,
|
| 896 |
+
"grad_norm": 2.234440803527832,
|
| 897 |
+
"learning_rate": 3.134920634920635e-05,
|
| 898 |
+
"loss": 0.0149,
|
| 899 |
+
"step": 1220
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 4.3811219946571685,
|
| 903 |
+
"grad_norm": 0.32285788655281067,
|
| 904 |
+
"learning_rate": 3.1150793650793656e-05,
|
| 905 |
+
"loss": 0.0119,
|
| 906 |
+
"step": 1230
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 4.416740872662511,
|
| 910 |
+
"grad_norm": 1.612191081047058,
|
| 911 |
+
"learning_rate": 3.095238095238095e-05,
|
| 912 |
+
"loss": 0.0173,
|
| 913 |
+
"step": 1240
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 4.452359750667854,
|
| 917 |
+
"grad_norm": 5.2718987464904785,
|
| 918 |
+
"learning_rate": 3.075396825396826e-05,
|
| 919 |
+
"loss": 0.0201,
|
| 920 |
+
"step": 1250
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 4.487978628673197,
|
| 924 |
+
"grad_norm": 0.8849789500236511,
|
| 925 |
+
"learning_rate": 3.055555555555556e-05,
|
| 926 |
+
"loss": 0.0144,
|
| 927 |
+
"step": 1260
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 4.52359750667854,
|
| 931 |
+
"grad_norm": 2.3857195377349854,
|
| 932 |
+
"learning_rate": 3.0357142857142857e-05,
|
| 933 |
+
"loss": 0.0202,
|
| 934 |
+
"step": 1270
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 4.559216384683882,
|
| 938 |
+
"grad_norm": 0.264274924993515,
|
| 939 |
+
"learning_rate": 3.0158730158730158e-05,
|
| 940 |
+
"loss": 0.0185,
|
| 941 |
+
"step": 1280
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 4.594835262689225,
|
| 945 |
+
"grad_norm": 2.823655605316162,
|
| 946 |
+
"learning_rate": 2.996031746031746e-05,
|
| 947 |
+
"loss": 0.0158,
|
| 948 |
+
"step": 1290
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 4.630454140694568,
|
| 952 |
+
"grad_norm": 1.4026707410812378,
|
| 953 |
+
"learning_rate": 2.9761904761904762e-05,
|
| 954 |
+
"loss": 0.018,
|
| 955 |
+
"step": 1300
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 4.666073018699911,
|
| 959 |
+
"grad_norm": 3.624246120452881,
|
| 960 |
+
"learning_rate": 2.9563492063492066e-05,
|
| 961 |
+
"loss": 0.0113,
|
| 962 |
+
"step": 1310
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 4.7016918967052534,
|
| 966 |
+
"grad_norm": 7.601430892944336,
|
| 967 |
+
"learning_rate": 2.9365079365079366e-05,
|
| 968 |
+
"loss": 0.016,
|
| 969 |
+
"step": 1320
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 4.737310774710597,
|
| 973 |
+
"grad_norm": 3.959538221359253,
|
| 974 |
+
"learning_rate": 2.916666666666667e-05,
|
| 975 |
+
"loss": 0.023,
|
| 976 |
+
"step": 1330
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 4.7729296527159395,
|
| 980 |
+
"grad_norm": 5.305376052856445,
|
| 981 |
+
"learning_rate": 2.8968253968253974e-05,
|
| 982 |
+
"loss": 0.0126,
|
| 983 |
+
"step": 1340
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 4.808548530721282,
|
| 987 |
+
"grad_norm": 3.621107816696167,
|
| 988 |
+
"learning_rate": 2.876984126984127e-05,
|
| 989 |
+
"loss": 0.0174,
|
| 990 |
+
"step": 1350
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 4.844167408726625,
|
| 994 |
+
"grad_norm": 1.0203227996826172,
|
| 995 |
+
"learning_rate": 2.857142857142857e-05,
|
| 996 |
+
"loss": 0.0152,
|
| 997 |
+
"step": 1360
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 4.879786286731968,
|
| 1001 |
+
"grad_norm": 1.6705145835876465,
|
| 1002 |
+
"learning_rate": 2.8373015873015875e-05,
|
| 1003 |
+
"loss": 0.0117,
|
| 1004 |
+
"step": 1370
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 4.915405164737311,
|
| 1008 |
+
"grad_norm": 0.6089171767234802,
|
| 1009 |
+
"learning_rate": 2.8174603174603175e-05,
|
| 1010 |
+
"loss": 0.0184,
|
| 1011 |
+
"step": 1380
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 4.951024042742653,
|
| 1015 |
+
"grad_norm": 1.959324836730957,
|
| 1016 |
+
"learning_rate": 2.797619047619048e-05,
|
| 1017 |
+
"loss": 0.0135,
|
| 1018 |
+
"step": 1390
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 4.986642920747997,
|
| 1022 |
+
"grad_norm": 1.1502999067306519,
|
| 1023 |
+
"learning_rate": 2.777777777777778e-05,
|
| 1024 |
+
"loss": 0.0218,
|
| 1025 |
+
"step": 1400
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 4.997328584149599,
|
| 1029 |
+
"eval_accuracy": 0.986584279670452,
|
| 1030 |
+
"eval_loss": 0.05108034238219261,
|
| 1031 |
+
"eval_runtime": 172.0304,
|
| 1032 |
+
"eval_samples_per_second": 104.423,
|
| 1033 |
+
"eval_steps_per_second": 1.633,
|
| 1034 |
+
"step": 1403
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 5.022261798753339,
|
| 1038 |
+
"grad_norm": 1.9193094968795776,
|
| 1039 |
+
"learning_rate": 2.7579365079365083e-05,
|
| 1040 |
+
"loss": 0.0109,
|
| 1041 |
+
"step": 1410
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 5.057880676758682,
|
| 1045 |
+
"grad_norm": 0.6912887096405029,
|
| 1046 |
+
"learning_rate": 2.7380952380952383e-05,
|
| 1047 |
+
"loss": 0.0092,
|
| 1048 |
+
"step": 1420
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 5.0934995547640245,
|
| 1052 |
+
"grad_norm": 0.3243821859359741,
|
| 1053 |
+
"learning_rate": 2.718253968253968e-05,
|
| 1054 |
+
"loss": 0.0107,
|
| 1055 |
+
"step": 1430
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 5.129118432769368,
|
| 1059 |
+
"grad_norm": 1.4951286315917969,
|
| 1060 |
+
"learning_rate": 2.6984126984126984e-05,
|
| 1061 |
+
"loss": 0.0057,
|
| 1062 |
+
"step": 1440
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 5.1647373107747105,
|
| 1066 |
+
"grad_norm": 1.5525418519973755,
|
| 1067 |
+
"learning_rate": 2.6785714285714288e-05,
|
| 1068 |
+
"loss": 0.0084,
|
| 1069 |
+
"step": 1450
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 5.200356188780053,
|
| 1073 |
+
"grad_norm": 1.0544503927230835,
|
| 1074 |
+
"learning_rate": 2.6587301587301588e-05,
|
| 1075 |
+
"loss": 0.0143,
|
| 1076 |
+
"step": 1460
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 5.235975066785397,
|
| 1080 |
+
"grad_norm": 5.903578758239746,
|
| 1081 |
+
"learning_rate": 2.6388888888888892e-05,
|
| 1082 |
+
"loss": 0.0135,
|
| 1083 |
+
"step": 1470
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 5.271593944790739,
|
| 1087 |
+
"grad_norm": 0.9685418009757996,
|
| 1088 |
+
"learning_rate": 2.6190476190476192e-05,
|
| 1089 |
+
"loss": 0.0122,
|
| 1090 |
+
"step": 1480
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 5.307212822796082,
|
| 1094 |
+
"grad_norm": 1.8341890573501587,
|
| 1095 |
+
"learning_rate": 2.5992063492063496e-05,
|
| 1096 |
+
"loss": 0.0174,
|
| 1097 |
+
"step": 1490
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 5.342831700801424,
|
| 1101 |
+
"grad_norm": 1.3028405904769897,
|
| 1102 |
+
"learning_rate": 2.5793650793650796e-05,
|
| 1103 |
+
"loss": 0.0074,
|
| 1104 |
+
"step": 1500
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 5.378450578806768,
|
| 1108 |
+
"grad_norm": 0.8573488593101501,
|
| 1109 |
+
"learning_rate": 2.5595238095238093e-05,
|
| 1110 |
+
"loss": 0.0123,
|
| 1111 |
+
"step": 1510
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 5.41406945681211,
|
| 1115 |
+
"grad_norm": 0.8495379686355591,
|
| 1116 |
+
"learning_rate": 2.5396825396825397e-05,
|
| 1117 |
+
"loss": 0.0092,
|
| 1118 |
+
"step": 1520
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 5.449688334817453,
|
| 1122 |
+
"grad_norm": 4.806192874908447,
|
| 1123 |
+
"learning_rate": 2.5198412698412697e-05,
|
| 1124 |
+
"loss": 0.0117,
|
| 1125 |
+
"step": 1530
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 5.485307212822796,
|
| 1129 |
+
"grad_norm": 1.6827951669692993,
|
| 1130 |
+
"learning_rate": 2.5e-05,
|
| 1131 |
+
"loss": 0.0109,
|
| 1132 |
+
"step": 1540
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 5.520926090828139,
|
| 1136 |
+
"grad_norm": 1.2856072187423706,
|
| 1137 |
+
"learning_rate": 2.4801587301587305e-05,
|
| 1138 |
+
"loss": 0.0066,
|
| 1139 |
+
"step": 1550
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 5.556544968833482,
|
| 1143 |
+
"grad_norm": 1.8056079149246216,
|
| 1144 |
+
"learning_rate": 2.4603174603174602e-05,
|
| 1145 |
+
"loss": 0.0103,
|
| 1146 |
+
"step": 1560
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 5.592163846838824,
|
| 1150 |
+
"grad_norm": 1.3088467121124268,
|
| 1151 |
+
"learning_rate": 2.4404761904761906e-05,
|
| 1152 |
+
"loss": 0.0071,
|
| 1153 |
+
"step": 1570
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 5.627782724844168,
|
| 1157 |
+
"grad_norm": 0.6147578954696655,
|
| 1158 |
+
"learning_rate": 2.4206349206349206e-05,
|
| 1159 |
+
"loss": 0.0135,
|
| 1160 |
+
"step": 1580
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 5.66340160284951,
|
| 1164 |
+
"grad_norm": 3.193516969680786,
|
| 1165 |
+
"learning_rate": 2.400793650793651e-05,
|
| 1166 |
+
"loss": 0.0128,
|
| 1167 |
+
"step": 1590
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 5.699020480854853,
|
| 1171 |
+
"grad_norm": 1.0563770532608032,
|
| 1172 |
+
"learning_rate": 2.380952380952381e-05,
|
| 1173 |
+
"loss": 0.0116,
|
| 1174 |
+
"step": 1600
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 5.734639358860196,
|
| 1178 |
+
"grad_norm": 4.569545269012451,
|
| 1179 |
+
"learning_rate": 2.361111111111111e-05,
|
| 1180 |
+
"loss": 0.0134,
|
| 1181 |
+
"step": 1610
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 5.770258236865539,
|
| 1185 |
+
"grad_norm": 1.7688589096069336,
|
| 1186 |
+
"learning_rate": 2.3412698412698414e-05,
|
| 1187 |
+
"loss": 0.0104,
|
| 1188 |
+
"step": 1620
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 5.805877114870881,
|
| 1192 |
+
"grad_norm": 0.6101610064506531,
|
| 1193 |
+
"learning_rate": 2.3214285714285715e-05,
|
| 1194 |
+
"loss": 0.0124,
|
| 1195 |
+
"step": 1630
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 5.841495992876224,
|
| 1199 |
+
"grad_norm": 6.201462745666504,
|
| 1200 |
+
"learning_rate": 2.3015873015873015e-05,
|
| 1201 |
+
"loss": 0.0154,
|
| 1202 |
+
"step": 1640
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 5.8771148708815675,
|
| 1206 |
+
"grad_norm": 0.349431574344635,
|
| 1207 |
+
"learning_rate": 2.281746031746032e-05,
|
| 1208 |
+
"loss": 0.0073,
|
| 1209 |
+
"step": 1650
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 5.91273374888691,
|
| 1213 |
+
"grad_norm": 2.9831745624542236,
|
| 1214 |
+
"learning_rate": 2.261904761904762e-05,
|
| 1215 |
+
"loss": 0.0116,
|
| 1216 |
+
"step": 1660
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 5.948352626892253,
|
| 1220 |
+
"grad_norm": 2.7858543395996094,
|
| 1221 |
+
"learning_rate": 2.2420634920634923e-05,
|
| 1222 |
+
"loss": 0.0099,
|
| 1223 |
+
"step": 1670
|
| 1224 |
+
},
|
| 1225 |
+
{
|
| 1226 |
+
"epoch": 5.983971504897596,
|
| 1227 |
+
"grad_norm": 1.6756843328475952,
|
| 1228 |
+
"learning_rate": 2.2222222222222223e-05,
|
| 1229 |
+
"loss": 0.0086,
|
| 1230 |
+
"step": 1680
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 5.998219056099733,
|
| 1234 |
+
"eval_accuracy": 0.9873079492317969,
|
| 1235 |
+
"eval_loss": 0.051540642976760864,
|
| 1236 |
+
"eval_runtime": 168.5925,
|
| 1237 |
+
"eval_samples_per_second": 106.553,
|
| 1238 |
+
"eval_steps_per_second": 1.667,
|
| 1239 |
+
"step": 1684
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 6.019590382902939,
|
| 1243 |
+
"grad_norm": 3.378427028656006,
|
| 1244 |
+
"learning_rate": 2.2023809523809524e-05,
|
| 1245 |
+
"loss": 0.0089,
|
| 1246 |
+
"step": 1690
|
| 1247 |
+
},
|
| 1248 |
+
{
|
| 1249 |
+
"epoch": 6.055209260908281,
|
| 1250 |
+
"grad_norm": 2.4920456409454346,
|
| 1251 |
+
"learning_rate": 2.1825396825396827e-05,
|
| 1252 |
+
"loss": 0.0096,
|
| 1253 |
+
"step": 1700
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 6.090828138913624,
|
| 1257 |
+
"grad_norm": 2.810772657394409,
|
| 1258 |
+
"learning_rate": 2.1626984126984128e-05,
|
| 1259 |
+
"loss": 0.0102,
|
| 1260 |
+
"step": 1710
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 6.126447016918967,
|
| 1264 |
+
"grad_norm": 0.348102331161499,
|
| 1265 |
+
"learning_rate": 2.1428571428571428e-05,
|
| 1266 |
+
"loss": 0.0146,
|
| 1267 |
+
"step": 1720
|
| 1268 |
+
},
|
| 1269 |
+
{
|
| 1270 |
+
"epoch": 6.16206589492431,
|
| 1271 |
+
"grad_norm": 1.5229803323745728,
|
| 1272 |
+
"learning_rate": 2.1230158730158732e-05,
|
| 1273 |
+
"loss": 0.0128,
|
| 1274 |
+
"step": 1730
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 6.1976847729296525,
|
| 1278 |
+
"grad_norm": 1.736290693283081,
|
| 1279 |
+
"learning_rate": 2.1031746031746032e-05,
|
| 1280 |
+
"loss": 0.011,
|
| 1281 |
+
"step": 1740
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 6.233303650934996,
|
| 1285 |
+
"grad_norm": 0.5413634777069092,
|
| 1286 |
+
"learning_rate": 2.0833333333333336e-05,
|
| 1287 |
+
"loss": 0.006,
|
| 1288 |
+
"step": 1750
|
| 1289 |
+
},
|
| 1290 |
+
{
|
| 1291 |
+
"epoch": 6.2689225289403385,
|
| 1292 |
+
"grad_norm": 1.1734927892684937,
|
| 1293 |
+
"learning_rate": 2.0634920634920636e-05,
|
| 1294 |
+
"loss": 0.0087,
|
| 1295 |
+
"step": 1760
|
| 1296 |
+
},
|
| 1297 |
+
{
|
| 1298 |
+
"epoch": 6.304541406945681,
|
| 1299 |
+
"grad_norm": 0.21049003303050995,
|
| 1300 |
+
"learning_rate": 2.0436507936507937e-05,
|
| 1301 |
+
"loss": 0.0074,
|
| 1302 |
+
"step": 1770
|
| 1303 |
+
},
|
| 1304 |
+
{
|
| 1305 |
+
"epoch": 6.340160284951024,
|
| 1306 |
+
"grad_norm": 1.5480449199676514,
|
| 1307 |
+
"learning_rate": 2.023809523809524e-05,
|
| 1308 |
+
"loss": 0.0072,
|
| 1309 |
+
"step": 1780
|
| 1310 |
+
},
|
| 1311 |
+
{
|
| 1312 |
+
"epoch": 6.375779162956367,
|
| 1313 |
+
"grad_norm": 1.130774736404419,
|
| 1314 |
+
"learning_rate": 2.003968253968254e-05,
|
| 1315 |
+
"loss": 0.0082,
|
| 1316 |
+
"step": 1790
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 6.41139804096171,
|
| 1320 |
+
"grad_norm": 0.9111453890800476,
|
| 1321 |
+
"learning_rate": 1.984126984126984e-05,
|
| 1322 |
+
"loss": 0.0081,
|
| 1323 |
+
"step": 1800
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 6.447016918967052,
|
| 1327 |
+
"grad_norm": 0.12502439320087433,
|
| 1328 |
+
"learning_rate": 1.9642857142857145e-05,
|
| 1329 |
+
"loss": 0.0095,
|
| 1330 |
+
"step": 1810
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 6.482635796972396,
|
| 1334 |
+
"grad_norm": 0.832012414932251,
|
| 1335 |
+
"learning_rate": 1.9444444444444445e-05,
|
| 1336 |
+
"loss": 0.0042,
|
| 1337 |
+
"step": 1820
|
| 1338 |
+
},
|
| 1339 |
+
{
|
| 1340 |
+
"epoch": 6.518254674977738,
|
| 1341 |
+
"grad_norm": 3.6857128143310547,
|
| 1342 |
+
"learning_rate": 1.924603174603175e-05,
|
| 1343 |
+
"loss": 0.0081,
|
| 1344 |
+
"step": 1830
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 6.553873552983081,
|
| 1348 |
+
"grad_norm": 2.5364391803741455,
|
| 1349 |
+
"learning_rate": 1.9047619047619046e-05,
|
| 1350 |
+
"loss": 0.0089,
|
| 1351 |
+
"step": 1840
|
| 1352 |
+
},
|
| 1353 |
+
{
|
| 1354 |
+
"epoch": 6.5894924309884235,
|
| 1355 |
+
"grad_norm": 23.89657211303711,
|
| 1356 |
+
"learning_rate": 1.884920634920635e-05,
|
| 1357 |
+
"loss": 0.0118,
|
| 1358 |
+
"step": 1850
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 6.625111308993767,
|
| 1362 |
+
"grad_norm": 2.562563180923462,
|
| 1363 |
+
"learning_rate": 1.8650793650793654e-05,
|
| 1364 |
+
"loss": 0.0096,
|
| 1365 |
+
"step": 1860
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"epoch": 6.66073018699911,
|
| 1369 |
+
"grad_norm": 3.3216705322265625,
|
| 1370 |
+
"learning_rate": 1.8452380952380954e-05,
|
| 1371 |
+
"loss": 0.0113,
|
| 1372 |
+
"step": 1870
|
| 1373 |
+
},
|
| 1374 |
+
{
|
| 1375 |
+
"epoch": 6.696349065004452,
|
| 1376 |
+
"grad_norm": 0.8016022443771362,
|
| 1377 |
+
"learning_rate": 1.8253968253968254e-05,
|
| 1378 |
+
"loss": 0.0054,
|
| 1379 |
+
"step": 1880
|
| 1380 |
+
},
|
| 1381 |
+
{
|
| 1382 |
+
"epoch": 6.731967943009796,
|
| 1383 |
+
"grad_norm": 3.4590423107147217,
|
| 1384 |
+
"learning_rate": 1.8055555555555555e-05,
|
| 1385 |
+
"loss": 0.0071,
|
| 1386 |
+
"step": 1890
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 6.767586821015138,
|
| 1390 |
+
"grad_norm": 0.12237533181905746,
|
| 1391 |
+
"learning_rate": 1.785714285714286e-05,
|
| 1392 |
+
"loss": 0.0073,
|
| 1393 |
+
"step": 1900
|
| 1394 |
+
},
|
| 1395 |
+
{
|
| 1396 |
+
"epoch": 6.803205699020481,
|
| 1397 |
+
"grad_norm": 1.500787615776062,
|
| 1398 |
+
"learning_rate": 1.7658730158730162e-05,
|
| 1399 |
+
"loss": 0.0055,
|
| 1400 |
+
"step": 1910
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 6.838824577025823,
|
| 1404 |
+
"grad_norm": 3.2452268600463867,
|
| 1405 |
+
"learning_rate": 1.746031746031746e-05,
|
| 1406 |
+
"loss": 0.0112,
|
| 1407 |
+
"step": 1920
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 6.874443455031167,
|
| 1411 |
+
"grad_norm": 0.4780968129634857,
|
| 1412 |
+
"learning_rate": 1.7261904761904763e-05,
|
| 1413 |
+
"loss": 0.0074,
|
| 1414 |
+
"step": 1930
|
| 1415 |
+
},
|
| 1416 |
+
{
|
| 1417 |
+
"epoch": 6.910062333036509,
|
| 1418 |
+
"grad_norm": 0.4808253347873688,
|
| 1419 |
+
"learning_rate": 1.7063492063492063e-05,
|
| 1420 |
+
"loss": 0.0096,
|
| 1421 |
+
"step": 1940
|
| 1422 |
+
},
|
| 1423 |
+
{
|
| 1424 |
+
"epoch": 6.945681211041852,
|
| 1425 |
+
"grad_norm": 1.5511194467544556,
|
| 1426 |
+
"learning_rate": 1.6865079365079367e-05,
|
| 1427 |
+
"loss": 0.0047,
|
| 1428 |
+
"step": 1950
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 6.9813000890471955,
|
| 1432 |
+
"grad_norm": 0.8193422555923462,
|
| 1433 |
+
"learning_rate": 1.6666666666666667e-05,
|
| 1434 |
+
"loss": 0.0057,
|
| 1435 |
+
"step": 1960
|
| 1436 |
+
},
|
| 1437 |
+
{
|
| 1438 |
+
"epoch": 6.999109528049867,
|
| 1439 |
+
"eval_accuracy": 0.9875306167891338,
|
| 1440 |
+
"eval_loss": 0.04618992656469345,
|
| 1441 |
+
"eval_runtime": 166.9392,
|
| 1442 |
+
"eval_samples_per_second": 107.608,
|
| 1443 |
+
"eval_steps_per_second": 1.683,
|
| 1444 |
+
"step": 1965
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 7.016918967052538,
|
| 1448 |
+
"grad_norm": 0.1917089819908142,
|
| 1449 |
+
"learning_rate": 1.6468253968253968e-05,
|
| 1450 |
+
"loss": 0.0042,
|
| 1451 |
+
"step": 1970
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 7.052537845057881,
|
| 1455 |
+
"grad_norm": 0.48844099044799805,
|
| 1456 |
+
"learning_rate": 1.626984126984127e-05,
|
| 1457 |
+
"loss": 0.0053,
|
| 1458 |
+
"step": 1980
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 7.088156723063223,
|
| 1462 |
+
"grad_norm": 1.94171941280365,
|
| 1463 |
+
"learning_rate": 1.6071428571428572e-05,
|
| 1464 |
+
"loss": 0.0064,
|
| 1465 |
+
"step": 1990
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 7.123775601068567,
|
| 1469 |
+
"grad_norm": 10.181970596313477,
|
| 1470 |
+
"learning_rate": 1.5873015873015872e-05,
|
| 1471 |
+
"loss": 0.0072,
|
| 1472 |
+
"step": 2000
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 7.159394479073909,
|
| 1476 |
+
"grad_norm": 0.4458209276199341,
|
| 1477 |
+
"learning_rate": 1.5674603174603176e-05,
|
| 1478 |
+
"loss": 0.0086,
|
| 1479 |
+
"step": 2010
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 7.195013357079252,
|
| 1483 |
+
"grad_norm": 1.125260829925537,
|
| 1484 |
+
"learning_rate": 1.5476190476190476e-05,
|
| 1485 |
+
"loss": 0.0064,
|
| 1486 |
+
"step": 2020
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 7.230632235084594,
|
| 1490 |
+
"grad_norm": 2.555405616760254,
|
| 1491 |
+
"learning_rate": 1.527777777777778e-05,
|
| 1492 |
+
"loss": 0.0047,
|
| 1493 |
+
"step": 2030
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 7.266251113089938,
|
| 1497 |
+
"grad_norm": 0.7010562419891357,
|
| 1498 |
+
"learning_rate": 1.5079365079365079e-05,
|
| 1499 |
+
"loss": 0.0027,
|
| 1500 |
+
"step": 2040
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 7.3018699910952805,
|
| 1504 |
+
"grad_norm": 0.19031189382076263,
|
| 1505 |
+
"learning_rate": 1.4880952380952381e-05,
|
| 1506 |
+
"loss": 0.004,
|
| 1507 |
+
"step": 2050
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 7.337488869100623,
|
| 1511 |
+
"grad_norm": 0.25679388642311096,
|
| 1512 |
+
"learning_rate": 1.4682539682539683e-05,
|
| 1513 |
+
"loss": 0.0024,
|
| 1514 |
+
"step": 2060
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 7.3731077471059665,
|
| 1518 |
+
"grad_norm": 0.22514384984970093,
|
| 1519 |
+
"learning_rate": 1.4484126984126987e-05,
|
| 1520 |
+
"loss": 0.0022,
|
| 1521 |
+
"step": 2070
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 7.408726625111309,
|
| 1525 |
+
"grad_norm": 1.700714111328125,
|
| 1526 |
+
"learning_rate": 1.4285714285714285e-05,
|
| 1527 |
+
"loss": 0.0069,
|
| 1528 |
+
"step": 2080
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 7.444345503116652,
|
| 1532 |
+
"grad_norm": 3.598724842071533,
|
| 1533 |
+
"learning_rate": 1.4087301587301587e-05,
|
| 1534 |
+
"loss": 0.0043,
|
| 1535 |
+
"step": 2090
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 7.479964381121994,
|
| 1539 |
+
"grad_norm": 0.9559303522109985,
|
| 1540 |
+
"learning_rate": 1.388888888888889e-05,
|
| 1541 |
+
"loss": 0.0069,
|
| 1542 |
+
"step": 2100
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 7.515583259127338,
|
| 1546 |
+
"grad_norm": 2.183023452758789,
|
| 1547 |
+
"learning_rate": 1.3690476190476192e-05,
|
| 1548 |
+
"loss": 0.0043,
|
| 1549 |
+
"step": 2110
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 7.55120213713268,
|
| 1553 |
+
"grad_norm": 0.6762595772743225,
|
| 1554 |
+
"learning_rate": 1.3492063492063492e-05,
|
| 1555 |
+
"loss": 0.0054,
|
| 1556 |
+
"step": 2120
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 7.586821015138023,
|
| 1560 |
+
"grad_norm": 0.12739060819149017,
|
| 1561 |
+
"learning_rate": 1.3293650793650794e-05,
|
| 1562 |
+
"loss": 0.0059,
|
| 1563 |
+
"step": 2130
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 7.622439893143366,
|
| 1567 |
+
"grad_norm": 0.8051192164421082,
|
| 1568 |
+
"learning_rate": 1.3095238095238096e-05,
|
| 1569 |
+
"loss": 0.0014,
|
| 1570 |
+
"step": 2140
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 7.658058771148709,
|
| 1574 |
+
"grad_norm": 1.1333496570587158,
|
| 1575 |
+
"learning_rate": 1.2896825396825398e-05,
|
| 1576 |
+
"loss": 0.0072,
|
| 1577 |
+
"step": 2150
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 7.6936776491540515,
|
| 1581 |
+
"grad_norm": 0.06419402360916138,
|
| 1582 |
+
"learning_rate": 1.2698412698412699e-05,
|
| 1583 |
+
"loss": 0.0044,
|
| 1584 |
+
"step": 2160
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 7.729296527159395,
|
| 1588 |
+
"grad_norm": 1.2567334175109863,
|
| 1589 |
+
"learning_rate": 1.25e-05,
|
| 1590 |
+
"loss": 0.0028,
|
| 1591 |
+
"step": 2170
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 7.7649154051647375,
|
| 1595 |
+
"grad_norm": 4.354968070983887,
|
| 1596 |
+
"learning_rate": 1.2301587301587301e-05,
|
| 1597 |
+
"loss": 0.0062,
|
| 1598 |
+
"step": 2180
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 7.80053428317008,
|
| 1602 |
+
"grad_norm": 0.17071713507175446,
|
| 1603 |
+
"learning_rate": 1.2103174603174603e-05,
|
| 1604 |
+
"loss": 0.0045,
|
| 1605 |
+
"step": 2190
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 7.836153161175423,
|
| 1609 |
+
"grad_norm": 0.057894542813301086,
|
| 1610 |
+
"learning_rate": 1.1904761904761905e-05,
|
| 1611 |
+
"loss": 0.0068,
|
| 1612 |
+
"step": 2200
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 7.871772039180766,
|
| 1616 |
+
"grad_norm": 4.474624156951904,
|
| 1617 |
+
"learning_rate": 1.1706349206349207e-05,
|
| 1618 |
+
"loss": 0.0043,
|
| 1619 |
+
"step": 2210
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 7.907390917186109,
|
| 1623 |
+
"grad_norm": 0.9991622567176819,
|
| 1624 |
+
"learning_rate": 1.1507936507936508e-05,
|
| 1625 |
+
"loss": 0.0074,
|
| 1626 |
+
"step": 2220
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 7.943009795191451,
|
| 1630 |
+
"grad_norm": 0.0822732001543045,
|
| 1631 |
+
"learning_rate": 1.130952380952381e-05,
|
| 1632 |
+
"loss": 0.0041,
|
| 1633 |
+
"step": 2230
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 7.978628673196794,
|
| 1637 |
+
"grad_norm": 1.9230424165725708,
|
| 1638 |
+
"learning_rate": 1.1111111111111112e-05,
|
| 1639 |
+
"loss": 0.0043,
|
| 1640 |
+
"step": 2240
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 8.0,
|
| 1644 |
+
"eval_accuracy": 0.9890892896904921,
|
| 1645 |
+
"eval_loss": 0.045326121151447296,
|
| 1646 |
+
"eval_runtime": 168.1961,
|
| 1647 |
+
"eval_samples_per_second": 106.804,
|
| 1648 |
+
"eval_steps_per_second": 1.671,
|
| 1649 |
+
"step": 2246
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 8.014247551202137,
|
| 1653 |
+
"grad_norm": 2.2943098545074463,
|
| 1654 |
+
"learning_rate": 1.0912698412698414e-05,
|
| 1655 |
+
"loss": 0.0047,
|
| 1656 |
+
"step": 2250
|
| 1657 |
+
},
|
| 1658 |
+
{
|
| 1659 |
+
"epoch": 8.04986642920748,
|
| 1660 |
+
"grad_norm": 0.5264036059379578,
|
| 1661 |
+
"learning_rate": 1.0714285714285714e-05,
|
| 1662 |
+
"loss": 0.0038,
|
| 1663 |
+
"step": 2260
|
| 1664 |
+
},
|
| 1665 |
+
{
|
| 1666 |
+
"epoch": 8.085485307212823,
|
| 1667 |
+
"grad_norm": 1.6306657791137695,
|
| 1668 |
+
"learning_rate": 1.0515873015873016e-05,
|
| 1669 |
+
"loss": 0.0023,
|
| 1670 |
+
"step": 2270
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 8.121104185218165,
|
| 1674 |
+
"grad_norm": 0.7527822852134705,
|
| 1675 |
+
"learning_rate": 1.0317460317460318e-05,
|
| 1676 |
+
"loss": 0.0034,
|
| 1677 |
+
"step": 2280
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 8.156723063223508,
|
| 1681 |
+
"grad_norm": 1.7613914012908936,
|
| 1682 |
+
"learning_rate": 1.011904761904762e-05,
|
| 1683 |
+
"loss": 0.0103,
|
| 1684 |
+
"step": 2290
|
| 1685 |
+
},
|
| 1686 |
+
{
|
| 1687 |
+
"epoch": 8.192341941228852,
|
| 1688 |
+
"grad_norm": 0.3057126998901367,
|
| 1689 |
+
"learning_rate": 9.92063492063492e-06,
|
| 1690 |
+
"loss": 0.002,
|
| 1691 |
+
"step": 2300
|
| 1692 |
+
},
|
| 1693 |
+
{
|
| 1694 |
+
"epoch": 8.227960819234195,
|
| 1695 |
+
"grad_norm": 0.1171383410692215,
|
| 1696 |
+
"learning_rate": 9.722222222222223e-06,
|
| 1697 |
+
"loss": 0.0038,
|
| 1698 |
+
"step": 2310
|
| 1699 |
+
},
|
| 1700 |
+
{
|
| 1701 |
+
"epoch": 8.263579697239537,
|
| 1702 |
+
"grad_norm": 0.17709870636463165,
|
| 1703 |
+
"learning_rate": 9.523809523809523e-06,
|
| 1704 |
+
"loss": 0.0104,
|
| 1705 |
+
"step": 2320
|
| 1706 |
+
},
|
| 1707 |
+
{
|
| 1708 |
+
"epoch": 8.29919857524488,
|
| 1709 |
+
"grad_norm": 2.5424516201019287,
|
| 1710 |
+
"learning_rate": 9.325396825396827e-06,
|
| 1711 |
+
"loss": 0.0028,
|
| 1712 |
+
"step": 2330
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 8.334817453250222,
|
| 1716 |
+
"grad_norm": 1.413736343383789,
|
| 1717 |
+
"learning_rate": 9.126984126984127e-06,
|
| 1718 |
+
"loss": 0.0019,
|
| 1719 |
+
"step": 2340
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"epoch": 8.370436331255565,
|
| 1723 |
+
"grad_norm": 0.8712714910507202,
|
| 1724 |
+
"learning_rate": 8.92857142857143e-06,
|
| 1725 |
+
"loss": 0.0044,
|
| 1726 |
+
"step": 2350
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 8.406055209260908,
|
| 1730 |
+
"grad_norm": 5.921164035797119,
|
| 1731 |
+
"learning_rate": 8.73015873015873e-06,
|
| 1732 |
+
"loss": 0.0027,
|
| 1733 |
+
"step": 2360
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 8.441674087266252,
|
| 1737 |
+
"grad_norm": 0.6600878238677979,
|
| 1738 |
+
"learning_rate": 8.531746031746032e-06,
|
| 1739 |
+
"loss": 0.0094,
|
| 1740 |
+
"step": 2370
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 8.477292965271594,
|
| 1744 |
+
"grad_norm": 0.046510376036167145,
|
| 1745 |
+
"learning_rate": 8.333333333333334e-06,
|
| 1746 |
+
"loss": 0.0073,
|
| 1747 |
+
"step": 2380
|
| 1748 |
+
},
|
| 1749 |
+
{
|
| 1750 |
+
"epoch": 8.512911843276937,
|
| 1751 |
+
"grad_norm": 0.2806013822555542,
|
| 1752 |
+
"learning_rate": 8.134920634920636e-06,
|
| 1753 |
+
"loss": 0.0074,
|
| 1754 |
+
"step": 2390
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 8.54853072128228,
|
| 1758 |
+
"grad_norm": 0.09230457246303558,
|
| 1759 |
+
"learning_rate": 7.936507936507936e-06,
|
| 1760 |
+
"loss": 0.0053,
|
| 1761 |
+
"step": 2400
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 8.584149599287622,
|
| 1765 |
+
"grad_norm": 1.893324851989746,
|
| 1766 |
+
"learning_rate": 7.738095238095238e-06,
|
| 1767 |
+
"loss": 0.006,
|
| 1768 |
+
"step": 2410
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 8.619768477292965,
|
| 1772 |
+
"grad_norm": 1.035352349281311,
|
| 1773 |
+
"learning_rate": 7.5396825396825394e-06,
|
| 1774 |
+
"loss": 0.0056,
|
| 1775 |
+
"step": 2420
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 8.655387355298307,
|
| 1779 |
+
"grad_norm": 2.519357204437256,
|
| 1780 |
+
"learning_rate": 7.3412698412698415e-06,
|
| 1781 |
+
"loss": 0.0019,
|
| 1782 |
+
"step": 2430
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 8.691006233303652,
|
| 1786 |
+
"grad_norm": 0.36915910243988037,
|
| 1787 |
+
"learning_rate": 7.142857142857143e-06,
|
| 1788 |
+
"loss": 0.0053,
|
| 1789 |
+
"step": 2440
|
| 1790 |
+
},
|
| 1791 |
+
{
|
| 1792 |
+
"epoch": 8.726625111308994,
|
| 1793 |
+
"grad_norm": 2.2446630001068115,
|
| 1794 |
+
"learning_rate": 6.944444444444445e-06,
|
| 1795 |
+
"loss": 0.0023,
|
| 1796 |
+
"step": 2450
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 8.762243989314337,
|
| 1800 |
+
"grad_norm": 0.3775590658187866,
|
| 1801 |
+
"learning_rate": 6.746031746031746e-06,
|
| 1802 |
+
"loss": 0.0039,
|
| 1803 |
+
"step": 2460
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 8.79786286731968,
|
| 1807 |
+
"grad_norm": 3.002919912338257,
|
| 1808 |
+
"learning_rate": 6.547619047619048e-06,
|
| 1809 |
+
"loss": 0.0047,
|
| 1810 |
+
"step": 2470
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 8.833481745325022,
|
| 1814 |
+
"grad_norm": 2.3306655883789062,
|
| 1815 |
+
"learning_rate": 6.349206349206349e-06,
|
| 1816 |
+
"loss": 0.0066,
|
| 1817 |
+
"step": 2480
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 8.869100623330365,
|
| 1821 |
+
"grad_norm": 0.08028317987918854,
|
| 1822 |
+
"learning_rate": 6.1507936507936505e-06,
|
| 1823 |
+
"loss": 0.0035,
|
| 1824 |
+
"step": 2490
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 8.904719501335707,
|
| 1828 |
+
"grad_norm": 1.2194794416427612,
|
| 1829 |
+
"learning_rate": 5.9523809523809525e-06,
|
| 1830 |
+
"loss": 0.0045,
|
| 1831 |
+
"step": 2500
|
| 1832 |
+
},
|
| 1833 |
+
{
|
| 1834 |
+
"epoch": 8.940338379341052,
|
| 1835 |
+
"grad_norm": 1.457444429397583,
|
| 1836 |
+
"learning_rate": 5.753968253968254e-06,
|
| 1837 |
+
"loss": 0.0042,
|
| 1838 |
+
"step": 2510
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 8.975957257346394,
|
| 1842 |
+
"grad_norm": 1.0034886598587036,
|
| 1843 |
+
"learning_rate": 5.555555555555556e-06,
|
| 1844 |
+
"loss": 0.0012,
|
| 1845 |
+
"step": 2520
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 8.9973285841496,
|
| 1849 |
+
"eval_accuracy": 0.988810955243821,
|
| 1850 |
+
"eval_loss": 0.04604629799723625,
|
| 1851 |
+
"eval_runtime": 167.0665,
|
| 1852 |
+
"eval_samples_per_second": 107.526,
|
| 1853 |
+
"eval_steps_per_second": 1.682,
|
| 1854 |
+
"step": 2526
|
| 1855 |
+
},
|
| 1856 |
+
{
|
| 1857 |
+
"epoch": 9.011576135351737,
|
| 1858 |
+
"grad_norm": 0.16471421718597412,
|
| 1859 |
+
"learning_rate": 5.357142857142857e-06,
|
| 1860 |
+
"loss": 0.0059,
|
| 1861 |
+
"step": 2530
|
| 1862 |
+
},
|
| 1863 |
+
{
|
| 1864 |
+
"epoch": 9.04719501335708,
|
| 1865 |
+
"grad_norm": 0.5944855213165283,
|
| 1866 |
+
"learning_rate": 5.158730158730159e-06,
|
| 1867 |
+
"loss": 0.0068,
|
| 1868 |
+
"step": 2540
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"epoch": 9.082813891362422,
|
| 1872 |
+
"grad_norm": 1.3986272811889648,
|
| 1873 |
+
"learning_rate": 4.96031746031746e-06,
|
| 1874 |
+
"loss": 0.0023,
|
| 1875 |
+
"step": 2550
|
| 1876 |
+
},
|
| 1877 |
+
{
|
| 1878 |
+
"epoch": 9.118432769367764,
|
| 1879 |
+
"grad_norm": 0.8911547660827637,
|
| 1880 |
+
"learning_rate": 4.7619047619047615e-06,
|
| 1881 |
+
"loss": 0.0023,
|
| 1882 |
+
"step": 2560
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 9.154051647373107,
|
| 1886 |
+
"grad_norm": 2.983407735824585,
|
| 1887 |
+
"learning_rate": 4.563492063492064e-06,
|
| 1888 |
+
"loss": 0.0039,
|
| 1889 |
+
"step": 2570
|
| 1890 |
+
},
|
| 1891 |
+
{
|
| 1892 |
+
"epoch": 9.189670525378451,
|
| 1893 |
+
"grad_norm": 0.6036033630371094,
|
| 1894 |
+
"learning_rate": 4.365079365079365e-06,
|
| 1895 |
+
"loss": 0.0065,
|
| 1896 |
+
"step": 2580
|
| 1897 |
+
},
|
| 1898 |
+
{
|
| 1899 |
+
"epoch": 9.225289403383794,
|
| 1900 |
+
"grad_norm": 9.958078384399414,
|
| 1901 |
+
"learning_rate": 4.166666666666667e-06,
|
| 1902 |
+
"loss": 0.0073,
|
| 1903 |
+
"step": 2590
|
| 1904 |
+
},
|
| 1905 |
+
{
|
| 1906 |
+
"epoch": 9.260908281389137,
|
| 1907 |
+
"grad_norm": 0.27652016282081604,
|
| 1908 |
+
"learning_rate": 3.968253968253968e-06,
|
| 1909 |
+
"loss": 0.003,
|
| 1910 |
+
"step": 2600
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 9.29652715939448,
|
| 1914 |
+
"grad_norm": 0.2213822901248932,
|
| 1915 |
+
"learning_rate": 3.7698412698412697e-06,
|
| 1916 |
+
"loss": 0.0015,
|
| 1917 |
+
"step": 2610
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"epoch": 9.332146037399822,
|
| 1921 |
+
"grad_norm": 0.18189971148967743,
|
| 1922 |
+
"learning_rate": 3.5714285714285714e-06,
|
| 1923 |
+
"loss": 0.0019,
|
| 1924 |
+
"step": 2620
|
| 1925 |
+
},
|
| 1926 |
+
{
|
| 1927 |
+
"epoch": 9.367764915405164,
|
| 1928 |
+
"grad_norm": 0.05522046610713005,
|
| 1929 |
+
"learning_rate": 3.373015873015873e-06,
|
| 1930 |
+
"loss": 0.001,
|
| 1931 |
+
"step": 2630
|
| 1932 |
+
},
|
| 1933 |
+
{
|
| 1934 |
+
"epoch": 9.403383793410507,
|
| 1935 |
+
"grad_norm": 2.0256989002227783,
|
| 1936 |
+
"learning_rate": 3.1746031746031746e-06,
|
| 1937 |
+
"loss": 0.0069,
|
| 1938 |
+
"step": 2640
|
| 1939 |
+
},
|
| 1940 |
+
{
|
| 1941 |
+
"epoch": 9.439002671415851,
|
| 1942 |
+
"grad_norm": 3.011732339859009,
|
| 1943 |
+
"learning_rate": 2.9761904761904763e-06,
|
| 1944 |
+
"loss": 0.0042,
|
| 1945 |
+
"step": 2650
|
| 1946 |
+
},
|
| 1947 |
+
{
|
| 1948 |
+
"epoch": 9.474621549421194,
|
| 1949 |
+
"grad_norm": 0.815857470035553,
|
| 1950 |
+
"learning_rate": 2.777777777777778e-06,
|
| 1951 |
+
"loss": 0.0021,
|
| 1952 |
+
"step": 2660
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 9.510240427426536,
|
| 1956 |
+
"grad_norm": 1.9958930015563965,
|
| 1957 |
+
"learning_rate": 2.5793650793650795e-06,
|
| 1958 |
+
"loss": 0.0008,
|
| 1959 |
+
"step": 2670
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 9.545859305431879,
|
| 1963 |
+
"grad_norm": 0.05693604424595833,
|
| 1964 |
+
"learning_rate": 2.3809523809523808e-06,
|
| 1965 |
+
"loss": 0.0028,
|
| 1966 |
+
"step": 2680
|
| 1967 |
+
},
|
| 1968 |
+
{
|
| 1969 |
+
"epoch": 9.581478183437222,
|
| 1970 |
+
"grad_norm": 4.779353618621826,
|
| 1971 |
+
"learning_rate": 2.1825396825396824e-06,
|
| 1972 |
+
"loss": 0.0071,
|
| 1973 |
+
"step": 2690
|
| 1974 |
+
},
|
| 1975 |
+
{
|
| 1976 |
+
"epoch": 9.617097061442564,
|
| 1977 |
+
"grad_norm": 0.09340893477201462,
|
| 1978 |
+
"learning_rate": 1.984126984126984e-06,
|
| 1979 |
+
"loss": 0.0029,
|
| 1980 |
+
"step": 2700
|
| 1981 |
+
},
|
| 1982 |
+
{
|
| 1983 |
+
"epoch": 9.652715939447907,
|
| 1984 |
+
"grad_norm": 1.627168893814087,
|
| 1985 |
+
"learning_rate": 1.7857142857142857e-06,
|
| 1986 |
+
"loss": 0.0038,
|
| 1987 |
+
"step": 2710
|
| 1988 |
+
},
|
| 1989 |
+
{
|
| 1990 |
+
"epoch": 9.68833481745325,
|
| 1991 |
+
"grad_norm": 0.4794100224971771,
|
| 1992 |
+
"learning_rate": 1.5873015873015873e-06,
|
| 1993 |
+
"loss": 0.0033,
|
| 1994 |
+
"step": 2720
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 9.723953695458594,
|
| 1998 |
+
"grad_norm": 0.7291146516799927,
|
| 1999 |
+
"learning_rate": 1.388888888888889e-06,
|
| 2000 |
+
"loss": 0.0024,
|
| 2001 |
+
"step": 2730
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 9.759572573463936,
|
| 2005 |
+
"grad_norm": 0.5635161399841309,
|
| 2006 |
+
"learning_rate": 1.1904761904761904e-06,
|
| 2007 |
+
"loss": 0.0009,
|
| 2008 |
+
"step": 2740
|
| 2009 |
+
},
|
| 2010 |
+
{
|
| 2011 |
+
"epoch": 9.795191451469279,
|
| 2012 |
+
"grad_norm": 0.027575811371207237,
|
| 2013 |
+
"learning_rate": 9.92063492063492e-07,
|
| 2014 |
+
"loss": 0.0015,
|
| 2015 |
+
"step": 2750
|
| 2016 |
+
},
|
| 2017 |
+
{
|
| 2018 |
+
"epoch": 9.830810329474621,
|
| 2019 |
+
"grad_norm": 0.4068202078342438,
|
| 2020 |
+
"learning_rate": 7.936507936507937e-07,
|
| 2021 |
+
"loss": 0.0019,
|
| 2022 |
+
"step": 2760
|
| 2023 |
+
},
|
| 2024 |
+
{
|
| 2025 |
+
"epoch": 9.866429207479964,
|
| 2026 |
+
"grad_norm": 0.08091460168361664,
|
| 2027 |
+
"learning_rate": 5.952380952380952e-07,
|
| 2028 |
+
"loss": 0.0021,
|
| 2029 |
+
"step": 2770
|
| 2030 |
+
},
|
| 2031 |
+
{
|
| 2032 |
+
"epoch": 9.902048085485307,
|
| 2033 |
+
"grad_norm": 0.043514594435691833,
|
| 2034 |
+
"learning_rate": 3.9682539682539683e-07,
|
| 2035 |
+
"loss": 0.005,
|
| 2036 |
+
"step": 2780
|
| 2037 |
+
},
|
| 2038 |
+
{
|
| 2039 |
+
"epoch": 9.93766696349065,
|
| 2040 |
+
"grad_norm": 0.17870892584323883,
|
| 2041 |
+
"learning_rate": 1.9841269841269841e-07,
|
| 2042 |
+
"loss": 0.0024,
|
| 2043 |
+
"step": 2790
|
| 2044 |
+
},
|
| 2045 |
+
{
|
| 2046 |
+
"epoch": 9.973285841495994,
|
| 2047 |
+
"grad_norm": 0.13601838052272797,
|
| 2048 |
+
"learning_rate": 0.0,
|
| 2049 |
+
"loss": 0.0017,
|
| 2050 |
+
"step": 2800
|
| 2051 |
+
},
|
| 2052 |
+
{
|
| 2053 |
+
"epoch": 9.973285841495994,
|
| 2054 |
+
"eval_accuracy": 0.9900356268091739,
|
| 2055 |
+
"eval_loss": 0.04398982226848602,
|
| 2056 |
+
"eval_runtime": 167.3655,
|
| 2057 |
+
"eval_samples_per_second": 107.334,
|
| 2058 |
+
"eval_steps_per_second": 1.679,
|
| 2059 |
+
"step": 2800
|
| 2060 |
+
},
|
| 2061 |
+
{
|
| 2062 |
+
"epoch": 9.973285841495994,
|
| 2063 |
+
"step": 2800,
|
| 2064 |
+
"total_flos": 1.7813294264095949e+19,
|
| 2065 |
+
"train_loss": 0.08161249804892577,
|
| 2066 |
+
"train_runtime": 11582.5203,
|
| 2067 |
+
"train_samples_per_second": 62.035,
|
| 2068 |
+
"train_steps_per_second": 0.242
|
| 2069 |
+
}
|
| 2070 |
+
],
|
| 2071 |
+
"logging_steps": 10,
|
| 2072 |
+
"max_steps": 2800,
|
| 2073 |
+
"num_input_tokens_seen": 0,
|
| 2074 |
+
"num_train_epochs": 10,
|
| 2075 |
+
"save_steps": 500,
|
| 2076 |
+
"stateful_callbacks": {
|
| 2077 |
+
"TrainerControl": {
|
| 2078 |
+
"args": {
|
| 2079 |
+
"should_epoch_stop": false,
|
| 2080 |
+
"should_evaluate": false,
|
| 2081 |
+
"should_log": false,
|
| 2082 |
+
"should_save": true,
|
| 2083 |
+
"should_training_stop": true
|
| 2084 |
+
},
|
| 2085 |
+
"attributes": {}
|
| 2086 |
+
}
|
| 2087 |
+
},
|
| 2088 |
+
"total_flos": 1.7813294264095949e+19,
|
| 2089 |
+
"train_batch_size": 64,
|
| 2090 |
+
"trial_name": null,
|
| 2091 |
+
"trial_params": null
|
| 2092 |
+
}
|